Electronics Guide

Structure and Categories

The diagram consists of a horizontal spine representing the problem or effect, with angled bones branching off to represent major cause categories. For electronics manufacturing and reliability analysis, the traditional categories often include:

• Materials: Component quality, material specifications, incoming inspection results, storage conditions, and material compatibility issues
• Methods: Process procedures, work instructions, assembly sequences, testing protocols, and handling practices
• Machines: Equipment calibration, maintenance status, tooling wear, process capability, and equipment settings
• Measurements: Test accuracy, measurement uncertainty, calibration status, sampling plans, and inspection criteria
• Environment: Temperature, humidity, contamination, electrostatic discharge control, and cleanroom conditions
• People: Training adequacy, skill levels, work procedures adherence, fatigue, and communication effectiveness

Construction Process

Building an effective fishbone diagram requires systematic team participation. Begin by clearly defining the problem statement and placing it at the head of the fish. Draw the main spine and add the major category bones. Through brainstorming sessions, identify potential causes within each category and add them as smaller bones branching from the appropriate category. Continue subdividing causes into more specific factors until the team has exhausted all possibilities.

The completed diagram serves as a visual map of all potential causes, highlighting areas requiring further investigation. Teams can use voting or data analysis to prioritize which branches to investigate first, focusing resources on the most likely root causes.

Application Process

Begin with a clear problem statement, then ask why that problem occurred. Take the answer and ask why again. Continue this process, typically five times, until you reach a root cause that can be addressed with corrective action. The number five is a guideline rather than a rigid rule; some problems require fewer iterations, while complex failures may require more.

For example, investigating a field failure of a power supply might proceed as follows:

1. Why did the power supply fail? The output capacitor failed short circuit.
2. Why did the capacitor fail short circuit? The capacitor experienced overvoltage stress.
3. Why did overvoltage stress occur? Voltage spikes exceeded the capacitor's rating during load transients.
4. Why did voltage spikes exceed the rating? The capacitor voltage rating had inadequate design margin for worst-case transients.
5. Why was the design margin inadequate? The design review process did not include worst-case transient analysis.

This analysis reveals that the root cause is a gap in the design review process, not simply a defective capacitor. Corrective action should address the design review procedure to prevent similar issues in future products.

Best Practices

Effective application of the 5 Whys requires discipline and objectivity. Focus on processes and systems rather than assigning blame to individuals. Verify each answer with data or evidence before proceeding to the next why. When multiple branches emerge, follow each path to its conclusion. Document the entire chain of reasoning to support corrective action development and facilitate organizational learning.

Fault Tree Structure

Fault trees use standardized symbols to represent different types of events and logical relationships:

• Top Event: The undesired system-level failure being analyzed, typically shown as a rectangle at the top of the tree
• Intermediate Events: Failures or conditions that contribute to the top event, represented by rectangles
• Basic Events: Fundamental failures that cannot be further decomposed, shown as circles
• AND Gates: Logical operators indicating that all input events must occur for the output event to happen
• OR Gates: Logical operators indicating that any single input event can cause the output event
• Transfer Symbols: Triangles indicating continuation of the tree on another page or section

Construction Methodology

Begin by clearly defining the top event, ensuring it is specific and unambiguous. Working downward, identify the immediate causes of the top event and determine the logical relationship between them. Continue decomposing each intermediate event until reaching basic events that can be assigned probability values or are clearly identified as root causes.

For complex systems, fault trees can become large and intricate. Modularization techniques help manage complexity by identifying repeated subtrees that can be developed once and referenced multiple times. Computer-aided fault tree analysis tools facilitate construction, manipulation, and quantitative analysis of large fault trees.

Quantitative Analysis

When probability data is available for basic events, fault trees enable quantitative calculation of top event probability. Boolean algebra and minimum cut set analysis identify the smallest combinations of basic events that can cause the top event. This information guides prioritization of corrective actions and design improvements by highlighting the most significant contributors to system failure risk.

Event Tree Structure

An event tree begins with an initiating event on the left and progresses rightward through a series of branch points representing safety systems, operator actions, or other mitigating factors. At each branch point, the tree splits into success and failure paths. The rightmost column shows the possible end states, ranging from successful mitigation to various failure scenarios.

For electronics reliability analysis, initiating events might include component failures, environmental excursions, or human errors. Branch points could represent protective circuit activation, redundant system engagement, or maintenance intervention. End states describe the ultimate impact on system functionality, from continued operation to complete failure.

Integration with Fault Trees

Event tree analysis and fault tree analysis complement each other effectively. Fault trees can be developed for each branch point in an event tree to analyze the probability of success or failure at that point. This combination provides comprehensive analysis of both the causes of initiating events and the progression of consequences following those events.

Matrix Construction

Create a matrix with process inputs listed in rows and output characteristics in columns. Rate the importance of each output characteristic on a scale, typically 1 to 10. For each input-output combination, assign a correlation score indicating the strength of relationship between that input and output. Calculate priority scores by multiplying correlation scores by output importance ratings and summing across outputs for each input.

The resulting priority scores identify which process inputs most significantly affect critical outputs, guiding process control and improvement efforts toward the highest-leverage factors.

Application in Failure Analysis

During root cause analysis, cause and effect matrices help teams systematically evaluate potential causes against observed failure characteristics. By rating how well each potential cause explains each observed symptom, investigators can objectively prioritize which hypotheses to pursue with further investigation.

Constructing Pareto Charts

Gather failure data and categorize failures by type, location, symptom, or other relevant classification. Count occurrences in each category and sort categories in descending order. Create a bar chart showing category frequencies and overlay a cumulative percentage line. The resulting chart clearly shows which categories account for the majority of failures.

The classic Pareto principle suggests that roughly 80% of effects come from 20% of causes. While actual distributions vary, Pareto analysis consistently reveals that some failure modes dominate while others are relatively rare. Addressing the dominant modes first provides the greatest return on investigation and corrective action investment.

Stratified Analysis

Pareto analysis becomes more powerful when applied at multiple levels of stratification. After identifying the dominant failure category, analyze that category further to identify its dominant subcategories. Continue stratifying until reaching actionable root causes. This hierarchical approach efficiently directs investigation toward the most impactful findings.

Correlation Patterns

Examine scatter diagrams for patterns indicating correlation:

• Positive Correlation: Points cluster along an upward-sloping trend, indicating that as one variable increases, the other tends to increase
• Negative Correlation: Points cluster along a downward-sloping trend, indicating an inverse relationship
• No Correlation: Points scatter randomly with no discernible pattern
• Non-linear Relationships: Points follow curved patterns indicating more complex relationships

Correlation does not prove causation, but strong correlations warrant further investigation to determine whether a causal relationship exists. When combined with engineering knowledge and controlled experiments, scatter diagram analysis can provide compelling evidence for root cause identification.

Investigation Phases

A comprehensive failure investigation typically proceeds through defined phases:

1. Initial Response: Secure failed items, document initial conditions, notify stakeholders, and assess urgency
2. Information Gathering: Collect failure history, operating conditions, maintenance records, and similar failure reports
3. Non-destructive Examination: Conduct visual inspection, electrical testing, X-ray imaging, and other non-destructive techniques
4. Destructive Analysis: Perform cross-sectioning, decapsulation, and other techniques requiring sample alteration
5. Root Cause Determination: Synthesize findings to identify fundamental causes
6. Corrective Action Development: Define actions to prevent recurrence
7. Documentation and Communication: Prepare reports and share lessons learned

Investigation Planning

Before beginning detailed analysis, develop an investigation plan that outlines objectives, scope, team composition, timeline, and resource requirements. Consider what information and evidence are needed to support conclusions, and plan the sequence of activities to preserve options for subsequent analysis. Document the plan and obtain appropriate approvals before proceeding.

Physical Evidence Handling

When collecting failed components or assemblies, minimize handling to avoid introducing additional damage or contamination. Use appropriate packaging to protect against electrostatic discharge, mechanical shock, and environmental exposure. Label all items clearly with identification numbers, collection date, location, and collector's name. Maintain chain of custody logs documenting every transfer of physical evidence.

Documentation Evidence

Collect all relevant documentation including design specifications, manufacturing records, test data, maintenance logs, and operating procedures. Preserve electronic records in their native formats when possible. Document the source, date, and custodian for all collected documentation. Organize evidence systematically to facilitate analysis and support conclusions.

Witness Information

Interview personnel who observed the failure or have relevant knowledge of the circumstances. Conduct interviews as soon as practical after the failure while memories are fresh. Document interviews in writing, noting the interviewee, date, location, and key information provided. Distinguish between direct observations and opinions or interpretations.

Replication Planning

Design replication experiments to test specific root cause hypotheses. Define the conditions believed necessary to produce the failure, including environmental factors, electrical stresses, mechanical loads, and timing sequences. Identify observable indicators that would confirm successful replication. Plan measurements and data collection to capture relevant parameters during the experiment.

Accelerated Testing

When failures result from long-term degradation mechanisms, accelerated testing techniques can reproduce failures in practical timeframes. Apply elevated stresses such as temperature, voltage, humidity, or vibration to accelerate failure mechanisms. Ensure that acceleration factors are understood and that accelerated conditions do not introduce failure modes that would not occur under normal operating conditions.

Simulation and Modeling

When physical replication is impractical or impossible, computer simulation and modeling can support root cause analysis. Finite element analysis can predict mechanical stresses and thermal distributions. Circuit simulation can reproduce electrical transients and operating conditions. Validate simulation models against available physical data to establish confidence in their predictions.

Generating Hypotheses

Use brainstorming techniques and analytical tools such as fishbone diagrams to generate a comprehensive list of potential root causes. Consider all plausible explanations, even those that seem unlikely. Involve team members with diverse perspectives and expertise to avoid overlooking possibilities. Document all hypotheses for systematic evaluation.

Evaluating Hypotheses

Assess each hypothesis against available evidence. Consider whether the hypothesis explains all observed symptoms and facts. Identify predictions that would be true if the hypothesis is correct and test those predictions. Eliminate hypotheses that are inconsistent with established facts. Prioritize remaining hypotheses based on probability and testability.

Converging on Root Cause

Through iterative testing and refinement, narrow the field of hypotheses until one or a few best explain all evidence. The root cause should be specific enough to guide effective corrective action. If multiple root causes contributed to the failure, identify each and assess their relative importance. Document the logical chain connecting evidence to conclusions.

Implementation Verification

Confirm that corrective actions have been implemented as specified. Review documentation, inspect physical changes, and audit process modifications. Verify that personnel have been trained on new procedures. Document implementation status and any deviations from the original plan.

Effectiveness Verification

Demonstrate that implemented actions prevent recurrence of the original failure. Methods for effectiveness verification include:

• Testing: Subject corrected products or processes to conditions that previously caused failure
• Monitoring: Track failure rates over time to confirm sustained improvement
• Auditing: Periodically verify that process changes remain in place and effective
• Analysis: Review subsequent failures to confirm they are not related to the original root cause

Closure Criteria

Define objective criteria for closing corrective actions. Criteria should specify required evidence of implementation and effectiveness, including timeframes for monitoring. Obtain appropriate approvals before closing actions. Maintain records of verification activities and closure decisions for future reference.

Identifying Preventive Opportunities

Review root cause analysis findings to identify where similar conditions, designs, or processes exist elsewhere in the organization. Consider horizontal deployment of corrective actions to related products or production lines. Assess whether design standards, process specifications, or supplier requirements should be updated to prevent similar failures in future developments.

Risk-Based Prioritization

Prioritize preventive actions based on risk assessment. Consider the probability that similar failures could occur and the severity of consequences if they do. Focus resources on preventive actions with the greatest risk reduction benefit. Document risk assessments and prioritization decisions to support resource allocation.

Systemic Improvements

Look beyond individual product or process changes to identify opportunities for systemic improvements. Root cause analysis may reveal gaps in design review processes, supplier qualification procedures, testing protocols, or training programs. Addressing these systemic issues can prevent entire categories of failures rather than individual occurrences.

Capturing Lessons

Document lessons learned in a standardized format that captures essential information including:

• Background: Brief description of the failure event and its impact
• Root Cause: Summary of the fundamental cause identified through analysis
• Key Findings: Important discoveries made during the investigation
• Recommendations: Actions that should be taken to prevent similar failures
• Applicability: Products, processes, or situations where the lesson applies

Knowledge Management

Establish a repository for lessons learned that is searchable and accessible to relevant personnel. Organize lessons by product type, failure mode, technology area, or other relevant categories. Link lessons to related design standards, specifications, and procedures. Periodically review and update lessons to maintain accuracy and relevance.

Promoting Application

Make lessons learned an active part of engineering and manufacturing processes. Include lessons learned reviews in design review checklists. Reference relevant lessons in failure mode and effects analyses. Share significant lessons through technical bulletins, training sessions, and engineering forums. Measure and track the application of lessons learned to demonstrate value and identify improvement opportunities.